One of the most important material configurations in the field of microelectronics, thin film technology and surface coatings technology consists of a polymeric coating material deposited as a thin film onto a substrate of another material. The proper selection of polymeric materials and their on-going inspection for process control often requires very precise knowledge of the mechanical properties. Currently, mechanical analysis is destructive and a bulk property analysis. For many parts, coatings and thin films made from plastics, the bulk properties are not the crucial property which needs to be monitored or controlled. It is the mechanical properties of the surface of the coating or molded material which is of crucial interest. Most mechanical characterization instruments and methods are unable to evaluate real surface of a material, but instead look at a shallow but still bulk response for the film or material. Typically, hardness as a function of temperature, especially relative to glass transition temperature (Tg) of the surface of the material or thin film is not directly or efficiently measured.
Two types of instruments have been developed to facilitate studies of thin films: the depth-sensing nanoindentor (Pethica et al. (1983) Philos. Mag. A 48:593) and the atomic force microscope (AFM) (Binnig et al. (1986) Phys. Rev. Lett 56:930). For depth-sensing nanoindentation a controlled, variable force is applied to a sample by the indenter and the resulting displacement of the indenter is measured. The resulting “load vs displacement” data can then be analyzed to obtain mechanical properties, such as hardness and elastic modulus using well established models. Using the AFM-based indentation techniques, measurements are displacement controlled—the sample is displaced against a cantilever indenter via a piezoelectric actuator—and forces are inferred from the measured deflection of the cantilever and its (nominally) known spring constant. Several different cantilever configurations and displacement detection schemes have been developed to obtain these measurements. AFM technology can be used to analyze a number of mechanical and thermomechanical properties of films including hardness, Tg or modulus.
Since the invention of the scanning tunneling electron microscope (STEM) and the related atomic force microscope (AFM) (U.S. Pat. No. 4,343,993), numerous advances have been made in devices and methods using this technology (see e.g., U.S. Pat. Nos. 5,729,026; 5,804,709; 5,808,302; 5,907,096; 5,920,837 and 5,920,837). IBM continues to improve components and methods to exploit this powerful new technology with improvements in tip fabrication, sensors and integration, making practical very novel arrays of AFM tips (see e.g., U.S. Pat. Nos. 6,085,580; RE37299; 6,249,747; 6,369,400; 6,647,766; 6,862,925 and 5,856,967). Efforts to employ an AFM tip array for storage of information have been pursued and a number of crucial inventions have resulted from that effort. Devices for ultra high-density, high speed data storage applications using thermomechanical writing and thermal readout in thin polymer films as storage media work on the basis of atomic force microscopy (AFM) and are described in a number of patents and in the following publications representing the state of the art (see e.g., Mamin et al. (1995) IBM J. Res. Dev. 39(6):681-700; Grochowski and Hoyt (1996) IEEE Tran. Magnetics 32(3):1850-1854; Reid et al. (1997) Micromech. Syst. 6(4):294-302; Mansuripur, The Pres Syn. University of Cambridge, 1st Paperback Ed., pp. 463-464 and 575, 1998; Binnig et al. (1999) Applied Physics Let. 74(9):1339-1341; and Despont et al., “VLSI-NEMS Chip for AFM Data Storage”, Micro Electro Mech Sys., MEMS '99 IEEE Inter Conf. January 1999, pp. 564-569, 1999. In the process of pursuing AFM for information storage in a polymeric medium, it became evident that temperature and load profiles, made by AFM tips on potentially useful polymer materials, were providing information about the surface of the polymer film not previously measurable with such precision and relevance.
The basic components of an AFM, as illustrated in FIG. 1A, are the cantilever beam (2) and support (5), the probe tip (1), the specimen support (4) on which the specimen (3) is mounted, the x, y, z positioning stage (6), a light source for optical deflection sensing (7), a position sensitive detector (8) and control electronics for position staging (9). A key element of the AFM is its microscopic force sensor, or cantilever. The cantilever (2) is typically formed by one or more beams of silicon or silicon nitride, which are 100 to 500 μm long and about 0.5 to 5 μm thick. Mounted on the end of the cantilever is a sharp tip (1), also referred to herein as the indenter, that is used to sense a force between the sample and the tip. For normal topographic imaging the probe tip is also comprised of plain silicon or silicon nitride. In order to obtain information on mechanical properties such as hardness typically the tip is comprised of silicon, silicon nitride or diamond. Silicon tips are preferable because they are the easiest to microfabricate and they allow the resistive heater to be included. However, heatable silicon nitride and diamond probes have also been used.
The shape of the tip, in particular the radius of curvature at the apex, is one of the most critical parameters in materials testing by means of nanoindentation. The tip radius determines the scale factor relating the experiment loading force to the mechanically relevant local stress (see K. L. Johnson, Contact Mechanics, Cambridge University Press, 1985). Using an indenter with an apex radius of 10 nm, local stresses exceeding 10 GPa can be produced on hard specimens (elastic modulus on the order of 10 GPa) with loading forces on the order of 1 μN. As a result of the high stress acting on the indenter tip the latter wears rapidly resulting in a blunted apex after only a few indentation experiments. Materials hardness, also expressed in terms of yield stress, is assessed by measuring the stress value at which the material is irreversibly deformed. Hard plastics exhibit yield stresses on the order of several GPa below the glass transition temperature. Hence, tip-wear is a serious issue in such investigations.
While AFM, as well as, other currently existing techniques can be used to analyze the mechanical properties of the majority of thin films there are distinct disadvantages associated with these techniques which limit their general applicability. In particular, many current methods are lacking in their ability to provide information on Tg and hardness for many thin films for a number of reasons. First, in some instances the material can only be made in thin film form without sacrificing essential properties. An example is paints or resists which have to be cured under defined conditions, i.e. in the thin film state. Most current techniques however, require more material than a mere film. For example, for calorimetry approximately a milligram of material is needed. Thus, often the film is too thin for conventional hardness testing. Attempts to alleviate this problem using sharper tips and shallower indents as in “nanoindentation” suffer from rapid tip wear after only few indentations. With respect to glass transition temperature (Tg) the film is often not thick enough to allow the application of bulk averaging techniques such as Differential Scanning Calorimetry (DSC) or Dynamic Mechanical Thermal Analysis (DMTA), etc; or the film is made on a sample substrate which itself degrades thermally or interferes otherwise with the measurement. In particular, the sample may not be heatable to sufficiently high temperatures; and other, non-averaging techniques for surface characterization (such as AFM friction measurements) are often too tedious. Finally, occasionally the material may only be made as one of several phases of a compound for which techniques for the measurement of Tg and hardness are not yet available. Published methods are often not sufficiently reliable and their application is typically too tedious to be of practical use.
There remains a need for an improved method for measuring the mechanical properties of thin films which meet the following criteria: spatially resolves data on hardness and Tg in the sub 100 nm range (laterally), can be applied to thin films having a thickness of less than or equal to 10 nm; thereby only modifying the material by heat or plastic deformation within the same dimensions of a few nm or less (typically the full sample, usually a wafer of cm-dimensions or greater must be heated); allows for fast throughput of samples; alleviates the tip-wear problem associated with hardness testing, preferably allows for measurement of the most important properties in one measurement, i.e. Tg, hardness and roughness and more preferably thermomechanical shift-factors; and ideally can perform the measurements as a function of depth. Most preferably the improved method would allow for the in situ analysis by nondestructive means. This is particularly important for materials that are present on a delicate device.
Accordingly, an object of this invention is to provide a method for analyzing the mechanical properties of thin films that meets the criteria set forth above.